This is a work in progess, most ideas aren’t well fleshed out and will probably have to reorder it at some point to mantain some sort of narrative.
Working on this has been one of my favourite ways of procastinating from more important matters, so I will be updating it slowly over time.
Table of Contents
- Introduction
- Definitions
- History
- Philosophy
- Physics
- Evolution
- Intelligence
- Artificial-Intelligence
- Embodied Cognition
- Neuroscience
- Altered States of Consciousness
- Buddhism and Mindfulness
- Free Will
- Conclusions
Introduction
I want this post to act as a repository of research and ideas related to the hard problem of consciousness:
Why and how does an assembly of neurons , no matter how complex, give rise to perceptions and feelings that are consciously experienced.
This is a fundamentally different problem from others found in science, and humanity has been, in one way or another, dancing around this question for the longest time. Here I will compile theories and thoughts that are compelling to me, thus I will generally avoid religious explanations (thought Buddhism and Mindfulness will get their own chapter for reasons explained there). While the exploration of consciousness was until few decades ago considered a exclusively philosophical problem, the interest has grown in the scientific community over the years, mostly out of advancements in neuroscience research and new technologies that allow us to get a much closer look at the brain.
As it usually happens, the most interesting topics tend to be heavily interdisciplinary, meaning that you can approach a solution from many different fields, and to get really close requires a close collaboration between starkly different specialties (which was really difficult until recently). Those fields include (but are not limited to): neuroscience, physics, computer science, biology, philosophy, linguistics, psychology… One good thing out of this is that, if you can reach the same conclusion from different starting points, it’s very possible that there is some truth in there.
The intention is for the post to grow over time, and try to share the most recent developments I can find. I don’t think a blog is the best format for it, I’m was thinking of something more like a digital garden, but while I build that this will suffice. Also I don’t really like dividing the chapters the way I did since this isn’t a subject you can approach linearly.
The research of consciousness is a extremely rigorous field but still the recipient of much speculation. Don’t trust me in any of the things I say or quote (applicable to all of my blog) and do your own research. I have a tendency to believe in things that match well with my world model but I am wrong most of the time.
Finally, I’d like to preface that humans are especial in many ways, but being conscious is not one of them. Many of the things discussed here apply also to some animals, particularly mammals and octopus. That’s why I believe that animal suffering is so dissonant with our current (western) values, and we are mostly ignoring the problem.
Definitions
I will be introducing some concepts that will be used throughout the post. Without a doubt the main problem that I encountered in my research was the extreme conflation of words As many of this concepts have different interpretations depending on the field that they are being tackled from, I think it’s important spend some time and make sure that the meanings are aligned.
Mind: The mind is what thinks, feels, perceives, imagines, remembers, and wills, encompassing the totality of mental phenomena. It includes both conscious and unconscious processes.
Consciousness: Subjective experience, awareness of internal and external existence.
- There is not an accepted operational definition so it can work as an adjective or substantive depending on context.
- Many claim to have solved it by using a really strict definition. There still is a lot of discussion around types and borders that will be explored later.
- It is mostly accepted that there are different types of consciousness, some of them seem much more difficult than others to attain.
Qualia: refers to instances of subjective, conscious experience. (the “redness” of the red).
Hard problem of consciousness: “the problem of explaining why and how we have qualia or phenomenal experiences. That is to say, it is the problem of why we have personal, first-person experiences, often described as experiences that feel “like something"".
- Humans mostly agree that conscious beings are qualitatively different from others, and deserve certain rights. But depending on the definition of consciousness you use that will hold different groups of beings (for some definitions, almost everything (citation needed)). This approach to ethics seems to be a dead end if it’s the case that the hard problem is not real, and there is not some magic ingredient.
Reductive Materialism/The Identity Theory of Mind / Physicalism: every kind of mental state, including consciousness, just is a corresponding kind of physical state.
Panpsychism: literally means that everything has a mind, that mentality is fundamental and ubiquitous in the natural world. Depending on the flavor even atoms might have some kind of consciousness.
History and Challenges
While many famous philosophers have tried to answer the questions around the conscious experience, I’m not particularly interested in those accounts at the moment. They are valuable in many aspects, but I’m more interested in concrete models with experimental research behind them, even when they have already been found wrong.
It could be argued, as many respected figures do, that humans are not capable of understanding it. This view is called cognitive closure and is exemplified by the paper What is i like to be a bat?, the main argument being that the subjective character of experience cannot be reduced to materialism. This is a similar objection to that posed by the limits of science, and while valid complaints, my intuition really hates what they imply.
For much time the mainstream view on consciousness was that of mind-body dualism. Usually associated with Descartes, it denotes that either the view that mental phenomena are non-physical or that the mind and body are distinct and separable. This view has been largely replaced by monism approaches such as physicalism or enactivism.
Almost all competing theories of consciousness, it seems to me, have been so vague, fluffy, and malleable that they can only aspire to wrongness.”
- Scott Aaronson
The main problem is that, the same way we don’t have a real definition of consciousness, we have no way to test for it. The most used example are p-zombies, that are beings that that act exactly like a conscious being, but is not conscious. As we have no way to test for this, it’s possible that everyone (except myself) is a p-zombie (this is a useful line of thought when dealing with charged personal issues, at least for a low emotional intelligence guy like me (not really please don’t do this) ).
Funnily enough, one of the main challenges here is proving which theories are correct and which aren’t. The ARC project has this mission, based on adversarial collaboration, to prove which theory is more right after making predictions on a empirical experiment. While an interesting approach, it’s extremely narrow in scope (only a couple of compatible theories can be challenged at a time) and difficult to run. And even after all the effort the results remain inconclusive. Other approaches include developing a set of criteria with which to judge all theories of consciousness, and something called IBE (Inference to the Best Explanation), that as far as I understand is just seeing which one fit best with what we know with many words.
This is complicated more by the fact that many competing theories do not even have the same explanatory targets, and much of the language between fields has different implications. It has also been proposed that humans are badly qualified to find an answer to this problem, I really liked the abstract of the paper Is There Something it’s Like to be a Garden Snail, that acts as a successor of the “What’s it like to be a bat” one.
The question “are garden snails conscious?” or equivalently “is there something it’s like to be a garden snail?” admits of three possible answers: yes, no, and denial that the question admits of a yes-or-no answer. All three answers have some antecedent plausibility, prior to the application of theories of consciousness. All three answers retain their plausibility after the application of theories of consciousness. This is because theories of consciousness, when applied to such a different species, are inevitably question-begging and rely crucially on dubious extrapolation from the introspections and verbal reports of a single species.
There are a myriad of publications researching this problem and as far as I can tell no prevailing voice. The main topics of debate include: - The distinction between phenomenal consciousness and access consciousness. - Whether consciousness is a continuous stream of percepts or it is _ discrete - The dichotomous vs. graded nature of consciousness_ - Meta-awareness and the relationship between the conscious and the self. - Balancing a perspective “from within” with a perspective “from without” on conscious experiences.
These are some of the questions I want to explore on this post, but you will rarely find me taking a position here as my understanding is still shallow in many regards. One approach I’ve found useful is outlined in It From Bit, Revisited:
Confusion is especially tricky when we’re dealing with metaphysics, the basic structure of reality, since these differences condition the words we use to describe our differences. When our words mean different things, it’s hard to hold a productive conversation. This factor is very much in play when discussing consciousness.
The approach I’ve found works best here is to put energy into distilling philosophical confusion into binary yes/no scenarios we can all agree on. To find X’s such that “either we live in a universe where X is exactly true, or we live in a universe where X is false.” (This heuristic is partly inspired by Wheeler’s brilliant “it from bit” frame, from which I’ve also shamelessly stolen my title.) The goal is to redirect energy away from dissipative high-dimensional confusion and channel it toward evaluating definite scenarios; binary questions where it’s clear some fact must be either true or not true, with no muddled middle.
(…)
In this frame, perhaps “solving consciousness” involves finding ~30 crucial binary questions about how things work — one billion possible universes we might live in — then through inference and experiment determining which one we do live in. I don’t know the real number, but I think it’s discrete and (surprisingly) finite.
Some of this questions include: - Can consciousness be formalized? (describe via mathematics) - ‘strong formalism’ implies that there is an exact formal structure latent in reality. - ‘weak formalism’ stops at math being instrumentally useful for our understanding. - Does consciousness have conservation laws? - Is there a Bekenstein bound on it? - Is there some kind of game theory of intelligent civilizations? - do we live in a universe with finite or infinite potential value? - The Katechon Hypothesis, by Anatoly Karlin - The Unz Review
Hopefully it’s clear now that most of this questions will remain unanswered for generations to come, and that they are intrinsically related to own own understanding of the universe
Models
In this case, I’m differentiation between models and theories of consciousness on the basis of reach and falsifiability. A model is falsifiable and it only attempts to explain the mechanisms of consciousness. Theories are typically much broader and abstract, and usually integrate many models.
Most scientific writing assumes that consciousness is an emergent property of brain activity (mainly because if it was any other way it would be impossible to study).
Integrated Information Theory (IIT)
-
Giulio Tononi - “An Information Integration Theory of Consciousness” (2004)
- Tononi presents the Integrated Information Theory (IIT), which attempts to quantify consciousness and explain its emergence based on the integration of information in the brain.
vaia OSF
Multiple Drafts Model (MDM)
Multiple Drafts Model: Introduced by Daniel Dennett in the book ‘Consciousness Explained’, argues for the brain being a “bundle of semi-independent agencies”, consciousness being the apparently serial account for the brain’s underlying process in which multiple calculations are happening at once. It denies the existence of Qualia and places emphasis on the parallel processing of information. What we experience as a coherent stream of consciousness is actually a post-hoc narrative created by our brain, that also helps to smooths over gaps and inconsistencies in our perception and cognition.
The name “multiple drafts” comes the notion that various interpretations of sensory input and cognitive processes occur simultaneously, and those are being constantly updated and competing for dominance. The key here is that our conscious experience is not a complete, moment-to-moment representation of the world., but a distributed phenomenon where it emerges from the interplay of these multiple drafts. Here consciousness is not a distinct “thing” but more like a way of describing the brain’s ongoing activity .
While this model is mostly a theoretical framework, some experiments have been made that support this theories:
- Change Blindness: refers to the failure to detect changes in visual scenes, often substantial ones, when the change occurs during a brief interruption or distraction. Recently it was discovered that this changes can also occur gradually. The idea is that, while the visual experience is continuously updated, not all updates reach conscious awareness. This is basically the role of attention, which we will later go into.
- Attentional blink: typically measured using rapid serial visual presentation, the second of two target stimuli presented in rapid succession (typically within 200-500 ms) is often missed. This gets into other of the quirks of the MDM, that the conscious experience of an event is spread out over time, not localized to a single moment, so there are complex temporal dynamics at play.
Mos of this experiments are not conclusive in any way, and it could be argued that one cannot look into the functioning of consciousness with this kind of sensory tricks, as most of this can be explained as shortcomings of our attention mechanism. Also many of this experiments can also be explained with other theories, particularly the Global Workspace Theory seems conciliable with this one.
Recurrent Processing Theory (RPT)
Higher-Order Theories (HOT)
Global Workspace Theory (GWT)
(GNWT) (4) claims that consciousness is instantiated by the global broadcasting and amplification of information across an interconnected network of prefrontal-parietal areas and many high-level sensory cortical areas.
https://en.wikipedia.org/wiki/Global_workspace_theory
Orchestrated Objective Reduction (Orch-OR) Theory
Predictive Coding/Perceptual Inference Models
- Key Idea: The brain constantly generates and updates a model of the environment based on sensory inputs. Consciousness arises from the brain’s attempts to minimize the difference between its predictions and actual sensory inputs.
- Key Proponents: Karl Friston
Phenomenal Consciousness vs. Access Consciousness
- Phenomenal Consciousness (P-consciousness): Refers to the qualitative, subjective aspects of conscious experience (what it feels like).
- Access Consciousness (A-consciousness): Refers to the aspects of consciousness that are available for reasoning, reporting, and control of behavior.
- Key Proponent: https://en.wikipedia.org/wiki/Ned_Block
https://en.wikipedia.org/wiki/Anil_Seth!!!!!!!!
https://www.astralcodexten.com/p/your-book-review-consciousness-and Dehaene’s Consciousness And The Brain a perception or a thought is conscious if you can report on it. For Dehaene, attention is the gating mechanism that decides which information is allowed to enter consciousness. attention is an unconcious process (maybe attention isnt all you need ?)
Dual-aspect monism
Taking Monism Seriously – Opentheory.net
Pete Mandik’s qualia quietism and meta-Illusionism
Qualia quietism is a view according to which the terms and concepts that typically feature a discussion of qualia or phenomenal states consist of a viciously circular set of mutually interdefining terms that have no clear or conceptually distinct content such that we can say anything meaningful about them at all. Essentially the very notion of qualia or phenomenal states is meaningless and there is nothing substantive to say about them.
Physics
Modern physics rest on the assumption that on really small scales (at and below the scale of atoms) particles exist on multiple positions at the same time. Decoherence happens when interaction with the environment forces a particle to commit to a specific position, and collapse.
The 2 main interpretations for this are the Copenhagen interpretation , that suggested that decoherence is random, and Hugh Everett’s many-worlds interpretation (MWI), the latter being the most widely accepted on recent years. While difficult to wrap your head around, it has found the most success with experimental results, and there is reason for thinking that it is linked to consciousness.
As I started researching this part, I was quickly confronted with a multitude of new problems that
quantum x.com
Entropy | Free Full-Text | Testing the Conjecture That Quantum Processes Create Conscious Experience
Can Quantum Physics Explain Consciousness After All? - YouTube
generally just physicists ignoring neuroscience.
https://youtu.be/0nOtLj8UYCw?si=SF0CuHm8GOuVyHb_ cool ideas 17:30 are we just matching 2 things we don’t understand together?
Evolution
“The evidence seems compelling, indeed overwhelming, that fundamental aspects of our mental and social life, including language, are determined as part of our biological endowment, not acquired by learning, still less by training, in the course of our experience. Many find this conclusion offensive. They would prefer to believe that humans are shaped by their environment, not that they develop in a manner that is predetermined in essential respects. ” Chomsky “Language and Problems of Knowledge: The Managua Lectures” 1988
While in the past this was a topic of heavy discussion, and many times heavily politicized, the behavioral view has lost most of it’s fuel and it’s now understood that humans come prepackaged with a bunch of knowledge. The main path through witch this information arrives is evolution, in the form of cognitive structures or predispositions.
Chomsky is one of the most famous proponents of this battle, defending that the ability to acquire language is innately specified in humans, and that it can’t be solely explained by environmental factors or learned behavior. This nativist view implies that consciousness and cognitive abilities are shaped by pre-existing mental structures (much like the architecture of a neural network changes the kind of things it can learn), and thus probably have been shaped by evolution in some way.
The first step that was presumably achieved via evolution was the transition to sentient beings, understood as those with the capacity to experience feelings and sensations. This ability to perceive stimuli provides obvious survival advantages, and one can image how developing consciousness can also allow your species to survive longer. This is the most widely accepted materialist theory on how consciousness came to be, gradually and slowly, on a genealogical tree shaped structure. I find it intuitive to think that there were no big jumps from nothing to something, but there is some caveats, as evolution does not work at a steady rates, but instead more like steps. This steps can be of different sizes, and in period of heavy evolutionary pressure they tend to be bigger.
Genetics is basically this, a code on our DNA that result in structures that grow and develop in response to the environment, but starting from this encoded “template” that are our genes. Many would be pretty disappointed if that was all there is, and while I don’t discard other possible sources (panpsychism, idealism, simulation…), evolution is a much more complex process than it might seem at first glance. This would be the neo-Darwinian view.
I’ve looked around for some alternatives but haven’t been convinced by any of them. Epigenics (heritable changes in gene expression that do not involve changes to the underlying DNA sequence) looks interesting but not that important on the long run (given that the environment isn’t static). The Gaia hypothesis (the Earth and its biological systems behave as a vast self-regulating organism) sounds so cool but at the same time feels like a religious cope of our lack of understanding of complex systems. Process Structuralism also seems plausible but they all pale under the might of natural selection.
General Theory of Neural Networks - Rob Leclerc
Many of the processes that keep us alive do not imply consciousness. Apart from the obvious ones (respiration, digestion…), unconscious perceptions can induce negative emotions, you can unconsciously distinguish faces and abstract categories like “object” and the attention mechanism is unconsciously controlled.
Memory is a interesting case, they are inextricably linked with our identities, and can be both conscious or unconscious.
Both identity and memory play crucial roles in the development of conciousness. In a way, it is simplifying the world into a highly compressed model.
Either consciousness is a useless epiphenomenon, or we evolved it as an extra function to solve certain tasks
dna is not source code https://x.com/picopaco17/status/1793083282827198585?s=46
Vedgie Net The theory of Darwin Machines developed by William Calvin proposes that the brain, like life, implements evolution to search a near infinite problem space.
Memory
As we will explore more later in the Embodied Cognition chapter, our specific anatomy leads to the kind of things that we can do.
Affordances: usage of an object based of how we can interact with it
Wilson also describes at least five main (abstract) categories that combine both sensory and motor skills (or sensorimotor functions). The first three are working memory, episodic memory, and implicit memory; the fourth is mental imagery, and finally, the fifth concerns reasoning and problem—solving.
Margaret Wilson considers the embodied cognition perspective as fundamentally an evolutionary one, viewing cognition as a set of abilities that built upon, and still reflects, the structure of physical bodies and how human brains evolved to manage those bodies.[23]
thanks to their bipedal gait, early humans did not need their ‘forepaws’ for locomotion, facilitating them to manipulate the environment with the aid of tools. (“grasping an idea”)
Learning How Did Consciousness Evolve? An Illustrated Guide | The MIT Press Reader
Biology
hunger Why Does Ozempic Cure All Diseases? - by Scott Alexander
this guy claims he got super rithym after cancer surgery, Alex Cohen (@alexcohendrumsnyc) • Instagram photos and videos
Dreams
This is one of those special cases where, even thought we are also quite in the dark on the details of the process, by virtue of creating an altered state for the mind it allows us to study it in a new light. It seems like most sleep phases are unconscious. External stimulation during REM phases usually does not provoke any response, however, the brain does react show signs of consciousness when the stimulus is directly implanted into the brain via magnetic stimulation (TMS). It’s possible that we are conscious in dreams but cut off from outside perception (a bit how the silver eggheads work).
rats dreamss Rats dream about their tasks during slow wave sleep | MIT News | Massachusetts Institute of Technology
https://x.com/yacinemtb/status/1797305044972282294?s=46
https://www.theseedsofscience.pub/p/exit-the-supersensorium
dreams against the overfitted brain!! https://www.sciencedirect.com/science/article/pii/S2666389921000647
The brain simulates actions and their consequences during REM sleep | bioRxiv
Vision
Gibson again, Computationalist perspectives, for example, consider perceptual objects as an unreliable source of information upon which the mind must do some sort of inference
- Binocular rivalry occurs if your two eyes are presented with different images. In this case, most of the time you don’t see a weird overlay of the two images, but instead your conscious perception flips between seeing either one or the other. It’s a bit like staring at ambiguous images, but more consistent. So at any point in time, you perceive one of the images consciously and the other unconsciously.
- Attentional blinking describes the effect that after a conscious perception, you are consciously blind for anything that you see in the next 200-300ms. Let’s say you watch a fast stream of digits, each only visible for 100ms, and occasionally the stream contains a letter instead of a number. You are supposed to detect the letters. When you see an “M”, this enters your consciousness, and you detect it. But if 300ms (three images later), there is another letter “S”, you will not see it consciously. Actually, you will be sure that there was no “S” in the stream.
not perfect → masking: If the image is shown for 30ms, then people do not consciously see it, while for 60ms they do. This works with almost 100% accuracy, and has become the main workhorse for consciousness studies.
Language
Did the human capacity for language evolve gradually? No, it suddenly appeared around 50,000 years ago after a freak gene mutation: (Language and Mind, third edition, pg, 183-184) LOL same as consciousness probably wasnt a hard takeoff
gwern.net/doc/psychology/linguistics/2024-fedorenko.pdf
development of language at as recently as 40,000 years ago.
The Language of Thought Hypothesis (Stanford Encyclopedia of Philosophy)
Regions of the brain are not predetermined, they just happen to be good at specific things. In the case of language, it was developed, so the parts of our brain that is used for it (Brokeys area and barnakis area? ) were not specifically created for that.
Chomsky: Language Acquisition Device, prepared for to learn language, that’s why he was searching for a universal grammar. Embodied cognition is against this.
George Lakoff, for example, holds that reasoning and language, arise from the nature of bodily experiences and, thus, even people’s own metaphors have bodily references.[21]
reading and numbers: “Reading in the Brain” and “The Number Sense”,
Emotions
Minsky outlines the book as follows:[4]
- “We are born with many mental resources.”
- “We learn from interacting with others.”
- “Emotions are different Ways to Think.”
- “We learn to think about our recent thoughts.”
- “We learn to think on multiple levels.”
- “We accumulate huge stores of commonsense knowledge.”
- “We switch among different Ways to Think.”
- “We find multiple ways to represent things.”
- “We build multiple models of ourselves.”
…
Eristics as a formal scientific theory
Intelligence
Deeply correlated to consciousness, intelligence is a spectrum, is conciousness also one or there is a thresthold and it’s all-or-none?
Intelligence as the length of the prediction horizon Mahault Albarracin - Cognitive Science - Machine Learning Street Talk (MLST) | Podcast on Spotify
Compression
Human Knowledge Compression Contest: Frequently Asked Questions & Answers Intelligence has many faces, like creativity, solving problems, pattern recognition, classification, learning, induction, deduction, building analogies, optimization, surviving in an environment, language processing, knowledge, and many more. A formal definition incorporating all or at least most aspect of intelligence is difficult but not impossible. Informally, intelligence is an agent’s ability to achieve goals in a wide range of environments. The key concepts for a formal definition are compression and utility maximization, the other aspects are emergent phenomena.
Re: Mahoney/Sampo: [agi] Marcus Hutter’s lossless compression of human knowledge prize
Rationale for a Large Text Compression Benchmark
Artificial-Intelligence
As discussed before, there is deep relationship between intelligence and consciousness, but I believe it’s not a prerequisite for AGI (Artificial General Intelligence). Here we encounter a similar problem, as there is not a firm definition or a test to measure it. Turing test’s fell for the Chinese Room argument some time ago, and while some claim that we have already attained it, there are some problems left to solve.
Agency is one of the main one’s, and that does seem more linked to this blog’s topic. The potential in this branch of research lies in the ability to complement current analytic approaches in neuroscience, that seem to fall short in many aspects. We will develop AGI before we understand our brain, we will know we understand a brain once we can build a brain.
The pragmatic side in my wants to leave all the philosophy quests behind and side with GeoHotz
I think it’s going to end up looking like birds and planes. Planes fly in a much more “rigid” way, and they are much simpler than birds. But for our purposes, planes are more useful, and nobody would deny that they are “flying machines”
Nobody will deny that our artificial life are “thinking machines.” Nobody will care if the machine can “love” in the same way nobody cares if a plane eats fish. We will care if we’ve built something that can do everything useful that a human can do. (…) I have a far better intuition for how the brain works from studying machine learning than I ever got from neuroscience or cognitive psychology. Just as building model planes taught me more about flight than birds.
And to those who say, but the metaphors of the age are always used to describe the brain. Aka in the 1800s it was a steampunk brain, in the 1900s it was a calculator, and in the 2000s it’s a computer. To them I say, watch us build a brain, watch us build life. No other era did, we will. Because we are starting to understand.
And you’ll forget about that Chinese room so fast, when this “machine life” is cooking and cleaning and driving you around and making you fall in love with it. Of course it’s in love with you too. I mean, you feel like it is, right? There’s no such thing as consciousness, we are all p-zombies, and philosophy is for high schoolers who just discovered smoking weed.
This is a fine point of view as long as you are fine knowing that we are all living in a world void of qualia. I personally hate the implications, though I understand that most of the time we are in fact acting unconsiously.
There is a bunch of people that are already working on the problem of jumpstarting AI towards consciousness. There also is a bunch of people warnings on th edangers of that
https://www.lesswrong.com/posts/diruo47z32eprenTg/my-computational-framework-for-the-brain
Computation in Physical Systems (Stanford Encyclopedia of Philosophy) Artificial Intelligence (Stanford Encyclopedia of Philosophy) https://en.wikipedia.org/wiki/Philosophy_of_artificial_intelligence
redirect.cs.umbc.edu/courses/471/papers/turing.pdf
there is a jumpp → Cognitive AI
A Paradigm for AI Consciousness - The Seeds of Science
Quantum computers aside, all the computational systems we have are Turing Machines, so if we want to know if computers can reach this level, we need to stablish if all physics can be explained by Turing level computations. It would be surprising to me if this is true, since bits and atoms work in vastly different domains.
This means that it will be impossible for software working under our current framework to attain consciousness, but some hope remains for a pure hardware implementation, thought that seems extremely difficult and many centuries ahead.
Uploaded Intelligence
pantheon serie, uploaded intelligence, computable conciousness dont be like Sapponsky or Dawkins quatum thought…
does intelligente imply cnociousnesssss
“all cognition is embodied, interactive, and embedded in dynamically changing environments”.[19]
Funny that we have been creating tests for checking when machines will be considered concious and they have been regularly breaking them… Eliza (1966) Chess (Deepmind) Turing Test ChatGPT (2002) As discussed before cartesian dualism is pretty much out of the table, so turing test dont make much sense (chinese room argument).
probably a spectrum, probably very much related to intelligence, a trehshold perhaps.
Embodied-Cognition
For some time now I’ve held the belief that the fastest way to AGI is to stick the most advanced neural net that we can to a physical robot and let him interact freely with his environment. The main push to start writing this post was writing an article for my uni on Embodied Evolution[¹], and after using way too many words to explain how it can help with the exploration of the consciousness problem, I decided to start writing this.
By using the term embodied we mean to highlight two points: first that cognition depends upon the kinds of experience that come from having a body with various sensorimotor capacities, and second, that these individual sensorimotor capacities are themselves embedded in a more encompassing biological, psychological and cultural context.
The quote is from the book “The Embodied Mind: Cognitive Science and Human Experience” (1991). This work argues, drawing heavily on Buddhist philosophy and phenomenology , that consciousness arises from the dynamic interaction between an organism and its environment.
by speaking of “consciousness” we end up leading ourselves by thinking that there is any sort of thing as consciousness separated from behavioral and linguistic understandings.[42]
Definition-Classification
Embodiment Thesis: Many features of cognition are embodied in that they are deeply dependent upon characteristics of the physical body of an agent, such that the agent’s beyond-the-brain body plays a significant causal role, or a physically constitutive role, in that agent’s cognitive processing. source
This run contrary to the traditional view the cognition is something that an isolated brain did, emphasizing that cognition emerges in the relationship between an agent and the affordances provided by the environment rather than in the brain alone.
By traditional I mean something like the Cartesian model (disembodied theory of mind), according to which all mental phenomena are non-physical. Cartesian dualism proposes that the mind is entirely distinct from the body and can be successfully explained and understood without reference to the body or to its processes. This is proven to be mostly wrong, but is the baseline from which our current notions of cognition are built.
Embodied cognition should be better seen “as a research program rather than a well-defined unified theory” since the classification doesn’t really have well defined borders, but here is my best attempt to summarize it:
- enactive approach:
- our actions are important, act of sensing, actively sample parts of the world, you need to know where to go sample information.
- Radical enactivism: you can dispense of representationalism, its all in the body (there is no one control), if the body is sufficiently tuned to the environment you don’t even need cognition.
- Gibson, we perceive only in the service of how we can act upon them. only seen in virtue of how it can be manipulated.
- consider biodynamics to understand cognition
- extended cognition
- Any Clark, much of our cognitive capacity resides outside out mind (phones!, has cognition stopped or is it extended, what we mean by cognition?, is it all in the head)
- Limits cognitive processing neither to the brain nor even to the body, it extends it outward into the agent’s world
- Offloading Cognition: counting with fingers, some tasks are too hard to do only in the head. gestures help us communicate. Note taking and even talking.
- Second brain. - situated cognition:
- emphasizes that this extension is not just a matter of including resources outside the head but stressing the role of probing and changing interactions with the agent’s world.
- Cognition is situated in that it is inherently dependent upon the cultural and social contexts within which it takes place.[11]
Re-centering:
The embodiment thesis is pretty much accepted nowadays, leading to the emergence of the 4E features of cognition (embodied, embedded, enacted, and extended cognition). Now, cognition is no longer thought of as being instantiated in or by a single organism, rather:
It assumes that cognition is shaped and structured by dynamic interactions between the brain, body, and both the physical and social environments.
free energy principle, exchange between internal and external states, across this boundary (sensation in one direction, actions in the other) Mathematically you can exchange them, which means that your action upon the world becomes the world’s way of perceiving you, and the world acts upon you trough your perception of the world circular causality
Frontiers | Free-Energy Minimization and the Dark-Room Problem
Neuroscience
Free Energy Principle Bayesian theories of consciousness: a review in search for a minimal unifying model | Neuroscience of Consciousness | Oxford Academic Predictive Processing Active Inference Framework Dr. Thomas Parr - Active Inference Book - Machine Learning Street Talk (MLST) | Podcast on Spotify
A conscious perception does not get weaker, it gets stronger as it travels. Dehaene compares unconscious perception with a wave that runs out at the shore, while a conscious perception is like an avalanche that gains momentum. After 400ms, the conscious avalanche has activated large parts of the brain, which Dehaene calls global ignition. Moreover, all the brain parts synchronize, and information flows from all parts into each other (Global Neuronal Workspace).
in a conscious moment, the brain creates a coherent worldview, which we may call “sampling”, “coordination”, or “creating a memory item”. it works like a memo, which doesn’t contain the details of all the reports but just a condensed summary- But in the brain there is no single president to make decisions. There are regions that we call “executive control regions”, but they are notoriously hard to pinpoint, and there is no region which is always involved in decisions
One big disadvantage is that it is really slow. (500ms!), for reference, imagine playing a game at 2 fps… the other disadvantage involves the fact that its an exclusive system, so it can not be parallelized. While a conscious perception is processed the activity blocks off other perceptions. It has been shown that there is some kind of buffer that allows for that activity to be stores until it can be processed.
systems neuroscience
“Neurophilosophy: Toward a Unified Science of the Mind-Brain” (1986)
- Churchland argues for a neurobiological approach to understanding consciousness, emphasizing the importance of brain science.
https://www.nature.com/articles/d41586-019-02207-1
Altered-States-Of-Consciousness
to understand consciousness, the first step is to understand how unconscious processing works. Some glimpses come from blindsight patients, who lose the ability to see consciously due to brain damage. This can affect their whole field of vision, or just one hemisphere, or even just specific forms like lines. But they remain able to unconsciously process what they see. They will automatically walk around objects in their way, even though they swear that they don’t see them. The effect can also be artificially produced in monkeys.
- With Transcranial Magnetic Stimulation (TMS) we can stimulate specific brain regions. By stimulating sensory areas, we can induce hallucinations, and we can either make them conscious or unconscious by regulating the strength of TMS. By interfering with the right region at the right time, we can also prevent real perceptions from entering consciousness.
Dehaene’s schizophrenia: (partially) lose conscious perceptions. in order to reach consciousness, a masked signal needs to last “much longer” for schizophrenic patients. I
illusion, visual, auditory
Split-Brain: What We Know Now and Why This is Important for Understanding Consciousness - PMC
Buddhism and Mindfulness
In the intro I mentioned that I would not be covering religious approaches to the problem of consciousness. The reason for that is that many times I feel like they dodge many important questions by recurring to the omnipotence of their god and our lowly status as humans. And while it might be a technically correct argument (citation needed) when developed correctly, it doesn’t convince me in the slightest. To be honest my reference point for this is Jordan Peterson and some other Christian authors, and I have very little knowledge of Islam and Hebrew traditions, so I admit that I have not given them a fair chance, perhaps they will get a chance in another blog post.
Buddhism to me feels like an exception to this, particularly the practice of mindfulness seems to be engaging with this problem head on, and they have been doing that for a much longer time that any scientific tradition I can think of. I am firm believer of the Lindy effect, so they get some credit for that. The other main reason for my interest in this topic is my mother, that has been actively practicing for a long time.
She started with Yoga and basic meditation practice (I did that too for some time when I was little and remain really grateful for it). With time she became interested in more “advanced” teachings, and practiced a variety of methods before settling on a branch of mindfulness called (). While at points it might look like a cult from and outsiders view, I’ve come to value it, and I’m the extremely skeptic of these things!. At this point I believe (and it’s kind of proven() ) that most people in the world would greatly benefit from it’s practice, and that it has the potential of leaving many psychologists without a job (which is a good thing!).
But I’m not here to discuss the benefits of this method (I would need to regularly practice it to do that), for now I’m mostly interested on the philosophy that surrounds it, especially in places where it intersects with the question of consciousness. Luckily there has been ample scientific research on this, and I have a bunch of books that my mother has been wanting me to read, so there is no lack of resources.
Myssore
astanga yoga
https://www.ncbi.nlm.nih.gov/pmc/articles/PMC9226608/ https://en.wikipedia.org/wiki/The_Universe_in_a_Single_Atom https://buddhiststudies.stanford.edu/publications/taboo-subjectivity-toward-new-science-consciousness
The Dark Knight of the Soul - The Atlantic
Conclusions
Consciousness is hard.
Footnotes
[¹] : This is the best overview of the field I found, this is mine (in spanish)
Apuntesss
https://dynomight.net/consciousness/
The Analysis of Mind, by Bertrand Russell
Consciousness As Recursive Reflections - by Scott Alexander - How subjective experiences relate to objective physics. - The study of consciousness has differentiated its subject matter into the parts that can be studied normally (like wakefulness and complicated information processing juggling multiple bits of information, which in humans seems to require consciousness) from the weird part where there’s no consensus even on what would be the right questions to ask: subjective experience, a.k.a. “phenomenal” consciousness. / qualia is the individual instaces of this. - Brain imaging tech such as functional magnetic resonance imaging (fMRI) does not show thoughts. - Neural oscillation - Wikipedia - … how thoughts work, what happens when they become recursive… - qualia arise out of neuronal information processing much like biology arises out of chemistry. When chemical reaction chains build each other, they can achieve self-replication. When neuronal activities reflect each other, they can achieve self-reflection. _Many processes that know each other become one process that knows itself - comments on missing the hard problem, still triying to solve the soft one, interesting approach at that.
Anna Riedl: The Historical Development of Cognitive Science - YouTube
- Is There Something It’s Like to Be a Garden Snail? on animalssss ABSTRACT. The question “are garden snails conscious?” or equivalently “is there something it’s like to be a garden snail?” admits of three pos- sible answers: yes, no, and denial that the question admits of a yes-or- no answer. All three answers have some antecedent plausibility, prior to the application of theories of consciousness. All three answers retain their plausibility after the application of theories of consciousness. This is because theories of consciousness, when applied to such a different species, are inevitably question-begging and rely crucially on dubious extrapolation from the introspe ctions and verbal reports of a single species.
bayesian bullshit on brain.
Modularity_of_mind/ Society of Mind
active inference, possibly goes in agency Active InferenceThe Free Energy Principle in Mind, Brain, and Behavior | Books Gateway | MIT Press
Your Book Review: How Language Began - Astral Codex Ten chomsky Edward Feser: Chomsky on consciousness preproggraming THE GHOST IN THE MACHINE - YouTube maybe idont like him anymore? Central to Chomsky’s conception of language is the idea that grammar reigns supreme, and that human brains have some specialized structure for learning and processing grammar. In the writing of Chomsky and others, this hypothetical component of our biological endowment is sometimes called the narrow faculty of language (FLN) origin in erectus? Semiotics and the Origin of Language in the Lower Palaeolithic | Journal of Archaeological Method and Theory
embodied rats x.com
https://www.bitsofwonder.co/p/a-revolution-in-biology !!!!
https://nonint.com/2024/06/03/general-intelligence-2024/
https://www.science.org/doi/10.1126/science.1218616 !! It has been argued that the technological capability to fully simulate the human brain on digital computers will exist within a decade. This is taken to imply that we will comprehend its functioning, eliminate all diseases, and “upload” ourselves to computers (1). Although such predictions excite the imagination, they are not based on a sound assessment of the complexity of living systems. Such systems are characterized by large numbers of highly heterogeneous components, be they genes, proteins, or cells. These components interact causally in myriad ways across a very large spectrum of space-time, from nanometers to meters and from microseconds to years. A complete understanding of these systems demands that a large fraction of these interactions be experimentally or computationally probed. This is very difficult.
Have to create a forward/generative model of my own body and capabilities, I can use that to make inferences of how you are using your body.
theory of mind
eye movements
perceptual categorization
p-zombies
Probably some lesswrong can fit here
adversarial collaboration rests on identifying the most diagnostic points of divergence between competing theories, https://www.science.org/doi/10.1126/science.abj3259
Information processing
It would be a mistake, however, to suppose that cognition consists simply of building maximally accurate representations of input information…the gaining of knowledge is a stepping stone to achieving the more immediate goal of guiding behavior in response to the system’s changing surroundings. Explaining the Computational Mind
Aha but I didn’t define cognition, on purpose…
Consciousness is just the perception of the self in discrete points https://youtu.be/YVoXxYSiOBI?si=UrH2ksRwQbEBTah4
-
“Epiphenomenal Qualia” (1982)
- In this paper, Jackson introduces the thought experiment of Mary the color scientist to argue that there are aspects of consciousness (qualia) that are not captured by physical explanations.
“Realistic Monism: Why Physicalism Entails Panpsychism” (2006)
- Strawson argues that physicalism (the view that everything is physical) logically leads to panpsychism (the view that all matter has a form of consciousness).
-
“The Conscious Mind: In Search of a Fundamental Theory” (1996)
- Another book by Chalmers, this work expands on his earlier papers and argues for a dual-aspect theory of information.
In a way it’s similar to the discussion around the origin of life, thought I have mostly abandoned that one as it gets a bit too technical on the chemistry side, one of my many weak sides.